Orbital data centers and AI-driven probes.

“StarCloud Launches the First Orbital AI Data Centers

In early November 2025, StarCloud, a Washington-based startup focused on space technology, successfully launched a satellite into low Earth orbit carrying a high-performance Nvidia H100 GPU. This GPU is one hundred times more powerful than any chip previously used in space.” (Welp Magazine, StarCloud Orbital Data Centers: The Biggest AI Breakthrough in Space)

Orbital data centers (ODC) are an interesting yet expensive concept. Those data centers are things that can answer. The data center’s huge energy problem. Those orbital data centers can use satellite swarms that communicate with each other using cloud-based architecture. Each of those satellites uses solar panels for energy supply. The same technology is used in drone swarms. And swarms of small probes that can someday research our solar system. The cloud-based architecture and swarm technology. 

Guarantees that the entity can operate even if one brick or satellite loses its control. There are huge problems. That this kind of plan faces. The orbital data centers require new satellites. All the time. When new microchips are coming. To marketing. But. On the other hand, the calculation power of those satellite swarms is easy. To enhance. By launching more data satellites into the swarm. That means those systems are easy.  But expensive to complete. 

The data satellite can be like all other satellites. Those satellites form the bricks of the entity. Called a cloud-based data center. When a satellite turns old. It just sends information that is stored in it to the satellite that replaces it. And then that satellite can be returned to Earth by using parachutes to the yard of the factory. This is the way. On how technology, like microcircuits, can be recycled. 

The ODCs, orbital data centers, are not impossible. They are made of existing technology. The main problem is financing. There is also a possibility. That some third-party nation tries to damage those data centers. 

“Artist's concept illustration of the Jupiter Icy Moons Orbiter, a cancelled unmanned NASA exploration craft.” (Wikipedia, Jupiter Icy Moons Orbiter)

But those orbital data centers. This can be the pilot experiment, which opens the path. To the entire solar system. If we think that we send probes. And make bases on Mars and beyond, we need technology that solves communication problems. The AI-driven robots. Can make repairs to the nuclear-powered probes. And. They can make the bases for crews. But those systems require. 

A data-center-sized computing capacity. The AI-controlled probe must have models for all kinds of situations. Here we must notice that all the probes. In some models, the main probe can travel through the solar system and drop the sub-probes to the planets. This model is the lightweight version of the study project called Daedalus. 

The attempt to make a model of an interstellar probe that should research Alpha Centauri and some other solar systems. The Daedalus itself could travel past the targeted solar system. And launch the sub-probes that enter the solar system. The Daedalus requires stable bases all around the solar system. 

That I listed. In this text, remain at the conceptual level. The JIMO (Jupiter Icy Moons Orbiter) was a study project that planned to send a probe to Jupiter in 2015. But that probe was never completed. The JIMO remains. As the name of the AI-controlled Jupiter-orbiters study project. The real name for those futuristic probes. Is. Some kind of ancient god. The probes that will travel to the outer solar system must be larger and more versatile than modern probes. They must have a very powerful computer system that can operate independently in the outer solar system. 

“Daedalus spacecraft concept” (Wikipedia, Project Daedalus)

This means those systems must be very independent. In visions, the system that controls JIMO, SIMO, NIMO (Jupiter, Saturn, Uranus, and Neptune icy moon orbiters).  And the Kuiper belt probes are driven by AI that is similar to ChatGPT. A radio signal takes hours to travel to those probes. So they must operate very independently. Those probes would be larger than the probes that NASA sends to Mars. They can operate nuclear-powered vehicles. Those operate on the surface of those icy moons. 

The drone swarms could sink through the ice. They can use lasers or microwaves to travel through the icy shell. Or the probe can use high-power microwaves or lasers to remove ice and send probes to the icy oceans of Jupiter’s Europa and Saturn’s Enceladus moons. Those drone swarms can also operate. In the atmosphere of the gas giants. Or in the atmosphere of Saturn’s Titan moon. Those large probes use nuclear-powered systems. That guarantees the very long-term operational time. Their mission will be to map the far side of the solar system. But also test the AI that can operate independently in surprising situations. 

https://arstechnica.com/space/2026/03/orbital-data-centers-part-1-theres-no-way-this-is-economically-viable-right/

https://globalcarbonfund.com/carbon-news/orbital-data-center-guide-everything-you-need-to-know-about-this-next-gen-space-computing-technology/

https://www.datacenterdynamics.com/en/news/4ig-weighs-100m-investment-into-axiom-space-as-part-of-orbital-data-center-plan/

https://www.odchq.com/

https://satnews.com/2026/03/17/the-rise-of-the-orbital-data-center-solving-the-space-data-bottleneck/

https://www.starcloud.com/

https://welpmagazine.com/starcloud-orbital-data-centers-ai-in-space/

https://en.wikipedia.org/wiki/Jupiter_Icy_Moons_Orbiter

https://en.wikipedia.org/wiki/Project_Daedalus